The Genetic Basis of Diurnal Preference in Drosophila Melanogaster 1 Mirko Pegoraro1,4, Laura MM Flavell1, Pamela Menegazzi2

Home , Laboratory rat

bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 The genetic basis of diurnal preference in Drosophila melanogaster

2 Mirko Pegoraro1,4, Laura M.M. Flavell1, Pamela Menegazzi2, Perrine Columbine1, Pauline 3 Dao1, Charlotte Helfrich-Förster2 and Eran Tauber1,3 4 5 1. Department of Genetics, University of Leicester, University Road, Leicester, LE1 7RH, UK 6 2. Neurobiology and Genetics, Biocenter, University of Würzburg, Germany 7 3. Department of Evolutionary and Environmental Biology, and Institute of Evolution, 8 University of Haifa, Haifa 3498838, Israel 9 4. School of Natural Science and Psychology, Liverpool John Moores University, L3 3AF, UK 10

11 12 Corresponding author: E. Tauber, Department of Evolutionary and Environmental Biology, 13 and Institute of Evolution, University of Haifa, Haifa 3498838, Israel; Tel:+97248288784 14 Email: [email protected] 15

17 Classification: Biological Sciences (Genetics)

20 Keywords: Artificial selection, circadian clock, diurnal preference, nocturnality,

21 Drosophila

24 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

25 Abstract

26 Most animals restrict their activity to a specific part of the day, being diurnal, nocturnal or 27 crepuscular. The genetic basis underlying diurnal preference is largely unknown. Under 28 laboratory conditions, Drosophila melanogaster is crepuscular, showing a bi-modal activity 29 profile. However, a survey of strains derived from wild populations indicated that high 30 variability among individuals exists, with diurnal and nocturnal flies being observed. Using a 31 highly diverse population, we have carried out an artificial selection experiment, selecting 32 flies with extreme diurnal or nocturnal preference. After 10 generations, we obtained highly 33 diurnal and nocturnal strains. We used whole-genome expression analysis to identify 34 differentially expressed genes in diurnal, nocturnal and crepuscular (control) flies. Other than 35 one circadian clock gene (pdp1), most differentially expressed genes were associated with 36 either clock output (pdf, to) or input (Rh3, Rh2, msn). This finding was congruent with 37 behavioural experiments indicating that both light masking and the circadian pacemaker are 38 involved in driving nocturnality. The diurnal and nocturnal selection strains provide us with a 39 unique opportunity to understand the genetic architecture of diurnal preference. 40 41 42

43 Significance

44 Most animals are active during specific times of the day, being either diurnal or nocturnal.

45 Although it is clear that diurnal preference is hardwired into a species' genes, the genetic

46 basis underlying this trait is largely unknown. While laboratory strains of Drosophila usually

47 exhibit a bi-modal activity pattern (i.e., peaks at dawn and dusk), we found that strains from

48 wild populations exhibit a broad range of phase preferences, including diurnal and nocturnal

49 flies. Using artificial selection, we have generated diurnal and nocturnal strains that allowed

50 us to test the role of the circadian clock in driving nocturnality, and to search for the genes

51 that are associated with diurnal preference.

2 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

54 Introduction

55 Although time is one of the most important dimensions that define the species ecological

56 niche, it is often a neglected research area (1). Most animal species exhibit locomotor activity

57 that is restricted to a defined part of the day, and this preference constitutes the species-

58 specific temporal niche. Selection for activity during a specific time of the day may be driven

59 by various factors, including preferred temperature or light intensity, food availability and

60 predation. The genetic basis for such phase preference is largely unknown and is the focus of

61 this study.

62 The fact that within phylogenetic groups, diurnality preference is usually similar (2)

63 alludes to an underlying genetic mechanism. The nocturnality of mammals, for example, was

64 explained by the nocturnal bottleneck hypothesis (2), which suggested that all mammals

65 descended from a nocturnal ancestor. Nocturnality and diurnality most likely evolved through

66 different physiological and molecular adaptations (3). Two plausible systems that have been

67 targetted for genetic adaptations driven by diurnal preference are the visual system and the

68 circadian clock, the endogenous pacemaker that drives daily rhythms. The visual system of

69 most mammals is dominated by rods, yet is missing several cone photoreceptors that are

70 present in other taxa where a nocturnal lifestyle is maintained (4).

71 Accumulating evidence suggests that diurnal preference within a species is far more

72 diverse than previously thought. Laboratory studies (1) have often focused on a single

73 representative wild-type strain and ignored the population and individual diversity within a

74 species. In addition, experimental conditions in the laboratory setting (particularly light and

75 temperature) often fail to simulate the high complexity that exists in natural conditions (1).

76 Furthermore, many species shift their phase preference upon changes in environmental

77 conditions. Such “temporal niche switching” is undoubtedly associated with considerable

78 plasticity that may lead to rapid changes in behaviour. For example, the spiny mouse (Acomys

3 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

79 cahirinus) and the golden spiny mouse (A. russatus) are two desert sympatric species that

80 split their habitat: the common spiny mouse is nocturnal, whereas the golden spiny mouse is

81 diurnal. However, in experiments where the golden spiny mouse is the only species present,

82 the mice immediately reverted to nocturnal behaviour (5).

83 While plasticity plays an important role in diurnal preference, there is some evidence

84 for a strong genetic component underlying the variability seen among individuals. For

85 instance, twin studies (6) found higher correlation of diurnal preference in monozygotic twins

86 than in dizygotic twins, with the estimated heritability being as high as 40%. In addition, a

87 few studies in humans have reported a significant association between polymorphisms in

88 circadian clock genes and ‘morningness–eveningness’ chronotypes, including a

89 polymorphism in the promoter region of the period3 gene (7).

90 Drosophila melanogaster is considered a crepuscular species that exhibits a bimodal

91 locomotor activity profile (in the laboratory), with peaks of activity arising just before dawn

92 and dusk. This pattern is highly plastic and the flies promptly respond to changes in day-

93 length or temperature simulating winter or summer. It has been shown that rises in

94 temperature or irradiance during the day drives the flies to nocturnality, whereas low

95 temperatures or irradiances result in a shift to more prominent diurnal behaviour (8, 9). Such

96 plasticity was also demonstrated in studies showing that flies switch to nocturnality under

97 moon light (10, 11) or in the presence of other socially interacting flies (12).

98 Some evidence also alludes to the genetic component of phase preference in

99 Drosophila. Sequence divergence in the period gene underlies the phase difference seen in

100 locomotor and sexual rhythms between D. melanogaster and D. pseudoobscura (13). Flies

101 also show natural variation in the timing of adult emergence (eclosion), with a robust

102 response to artificial selection for the early and late eclosion phases having been shown,

103 indicating that substantial genetic variation underlies this trait (14).

4 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

104 Further support for a genetic component to phase preference comes from our previous

105 studies of allelic variation in the circadian-dedicated photoreceptor cryptochrome (CRY),

106 where an association between a pervasive replacement SNP (L232H) and the phases of

107 locomotor activity and eclosion was revealed (15). Studies of null mutants of the Clock gene

108 (Clkjrk) revealed that such flies became preferentially nocturnal (16), and that this phase

109 switch is mediated by elevated CRY in a specific subset of clock neurons (17). In other

110 experiments, mis-expression of Clk resulted in light pulses evoking longer bouts of activity,

111 suggesting that Clk plays a clock-independent role that modulates the effect of light on

112 locomotion (18).

113 Here, using 272 natural population strains from 33 regions in Europe and Africa, we

114 have generated a highly diverse population whose progeny exhibited a broad range of phase

115 preferences, with both diurnal and nocturnal flies being counted. We exploited this

116 phenotypic variability to study the genetic architecture of diurnal preference and identify loci

117 important for this trait using artificial selection, selecting for diurnal and nocturnal flies.

118

119

120 Results

121 Artificial selection for diurnal preference

122 Flies showed a rapid and robust response to selection for phase preference. After 10 selection

123 cycles, we obtained highly diurnal (D) and nocturnal (N) strains. The two control strains (C)

124 showed intermediate (crepuscular) behaviour (Fig. 1). To quantify diurnal preference, we

125 defined the ND ratio, quantitatively comparing activity during a 12 h dark period and during

126 a 12 h light period. As early as after one cycle of selection, the ND ratios of N and D flies,

127 and as compared to the original (control) population, were significantly different (Fig. 1A,

5 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

128 B). After 10 generation of selection, the N and D populations were highly divergent (Fig. 1B,

129 SI Appendix, Table S1).

130 The estimated heritability h2 was higher for diurnality (37.1%) than for nocturnality,

131 (8.4%) reflecting the asymmetric response of the two populations (SI Appendix, Table S1).

132 We estimated heritability by parents-offspring regression (Fig. 1C, D). The narrow-sense

2 = 133 heritability was lower but significant (h 14% p<0.05; Fig. 1C). The heritability value was

2 = 134 slightly higher when ND ratios of mothers and daughters were regressed (h 16% p<0.05;

135 Fig. 1D) but was minute and insignificant in the case of father-son regression (SI Appendix,

2 = 136 Fig. S1; h 2.5% NS).

137

138 Effects of Nocturnal/Diurnal phenotypes on fitness

139 A possible mechanism driving the observed asymmetric response to selection is unequal

140 allele frequencies, whereby a slower response to selection is associated with increased fitness

141 (19). We, therefore, tested whether our selection protocol asymmetrically affected the fitness

142 of the N and D populations. After ~15 overlapping generations (5 months in a 12h:12h

143 light:dark (LD) ) from the end of the selection, we tested viability, fitness and egg-to-adult

144 developmental time of the selection and control populations. While the survivorship of males

145 from the three populations was similar (χ2= 1.6, df=2, p=0.46; Fig. 2A), we found significant

146 differences in females. N females lived significantly longer than D females, while C females

147 showed intermediate values (χ2=7.6, df=2, p<0.05; Fig. 2A). The progeny number of N

148 females was larger than that of D females, with C females showing intermediate values

149 (♂F1,18 =5.12, p<0.05; ♀F1,18 =5.09, p<0.05; Fig. 2B). Developmental time (egg-to-adult),

150 another determinant of fitness, did not differ significantly between the nocturnal/diurnal

151 populations (♂F2,27 =0.43, p=0.65, NS; ♀ F2,27 =0.27, p=0.76, NS; Fig. 2C,D).

152

6 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

153 Effects on circadian behaviour

154 Since the circadian system is a conceivable target for genetic adaptations that underlie diurnal

155 preference, we tested whether the circadian clock of the N and D strains were affected by the

156 Nocturnal/Diurnal artificial selection. Accordingly, we recorded the locomotor activity of the

157 selection lines following three generations after completion of the selection protocol, and

158 measured various parameters of circadian rhythmicity.

159 The phase of activity peak in the morning (MP) and in the evening (EP) differed

160 between the populations (SI Appendix, Fig. S2A). As expected, the MP of the N population

161 was significantly advanced, as compared to that seen in both the C and D populations. The

162 EP of the N population was significantly delayed, as compared to that noted in the two other

163 populations. Concomitantly, the sleep pattern was also altered (SI Appendix, S2B), with N

164 flies sleeping much more during the day than did the other populations. While the D flies

165 slept significantly more than C and N flies during the night, there was no difference between

166 N and C flies.

167 In contrast to the striking differences seen between the selection lines in LD

168 conditions, such differences were reduced in DD conditions (Fig. 3, SI Appendix, Fig. S2C).

169 The period of the free-run of activity (FRP) was longer in the C flies than in D flies, while the

170 difference between the D and N groups was only marginally significant (Fig. 3A). No

171 significant difference in FRP was found between N and C flies. The phases (φ) of the three

172 populations did not differ significantly (Fig. 3B). We also tested the response of the flies to

173 an early night (ZT15) light stimulus and found no significant differences between the delay

174 responses (Fig. 3C).

175

7 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

176 Circadian differences between Nocturnal/Diurnal isogenic strains

177 To facilitate genetic dissection of the nocturnal/diurnal preference, we generated nocturnal,

178 diurnal and control isogenic strains (D*, N* and C*; one of each) from the selected

179 populations. The strains were generated using a crossing scheme involving strains carrying

180 balancer chromosomes. The ND ratios of the isogenic lines resembled those of the progenitor

181 selection lines (SI Appendix, Fig. S3A).

182 The circadian behaviour of the isogenic lines differed, with the N* line having a

183 longer FRP than both the D* and C* lines (Fig. 4A). The locomotory acrophase of the N*

184 line was delayed by about 2 h, as compared to the D* line, and by 1.38 h, as compared to the

185 C* line (F2,342:6.01, p<0.01; Fig. 4B). In contrast, circadian photosensitivity seemed to be

186 similar among the lines, as their phase responses to a light pulse at ZT15 did not differ

187 (F2,359:1.93, p:0.15, NS; Fig. 4D). Since eclosion is regulated by the circadian clock (20, 21),

188 we also compared the eclosion phase of the isogenic strains. Under LD, the eclosion phase of

189 D* flies was delayed by ~2 h, as compared to both N* and C* flies, whereas s no difference

190 between N* and C* flies was detected (Fig. 4C).

191 The isogenic strains also differed in terms of their sleep pattern (SI Appendix, Fig.

192 S3B). N* flies slept 4 h more than did D* flies during the day and ~2.5 h more than did C*

193 flies (F11,1468 = 61.32, p<0.0001). During the night, the pattern was reversed, with D* flies

194 sleeping almost 5 h more than N* flies and about 2 h more than C* flies (F11,1472 =104.70,

195 p<0.0001).

196

197 Diurnal preference is partly driven by masking

198 We reasoned that light masking (i.e., the clock-independent inhibitory or stimulatory effect of

199 light on behaviour) could be instrumental in driving diurnal preference. We thus monitored

200 fly behaviour in DD to assess the impact of light masking. We noticed that when N* flies

8 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

201 were released in DD, their nocturnal activity was much reduced, whereas their activity during

202 the subjective day increased (Fig. 5A). Indeed, the behaviour of N* and D* flies in DD

203 became quite similar (Fig. 5A). Congruently, when we analysed the ND ratios of these flies

204 in LD and DD, we found that that both N* and C* flies became significantly more “diurnal”

205 when released into constant conditions (N*, p<0.0001; C*, p<0.001). In contrast, the ratios of

206 D* flies did not significantly change in DD (p =0.94, NS). This result suggests that nocturnal

207 behaviour is at least partially driven by a light-dependent repression of activity (i.e., a light

208 masking effect).

209

210 Correlates of the molecular clock

211 To investigate whether differences in diurnal preference correlated with a similar shift in the

212 molecular clock, we measured the intensity of nuclear PERIOD (PER) in key clock neurons

213 (Fig 6). The peak of PER signals in ventral neurons (LNv: 5th-sLNv, sLNv, lLNv) was

214 delayed in N* flies, as compared to the timing of such signals in D* fly 5th-sLNv, sLNv and

215 lLNv neurons. In N* and D* flies, the phases of such peaks in dorsal neurons (DN, including

216 the clusters LNd, DN1and DN2) were similar (Fig 6).

217 We also measured the expression of the Pigment Dispersing Factor (PDF) in LNv

218 projections (Fig. 7, SI Appendix, Fig. S4). The signal measured in N* flies was lower than

219 that measured in D* flies during the first part of the day (in particular at ZT3 and ZT7) yet

220 increased during the day-night transition at ZT11 and ZT13. There were no differences seen

221 during the rest of the night.

222

223 Global transcriptional differences between Nocturnal/Diurnal strains

224 To gain insight into the genetics of diurnal preference, we profiled gene expression in fly

225 heads of individuals from the D*, C* and N* isogenic lines by RNAseq. We tested for

9 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

226 differentially expressed genes (DEG) in all pairwise contrasts among the three strains at two

227 time points. We found 34 DEGs at both ZT0 and ZT12 (SI Appendix, Table S3). An

228 additional 19 DEG were unique to ZT0 and 87 DEG were unique to ZT12 (SI Appendix,

229 Table S3). Functional annotation analysis (DAVID, https://david.ncifcrf.gov/ (22, 23))

230 revealed similarly enriched categories at ZT0 and ZT12 (Fig. 8). The predicted products of

231 the DEGs were largely assigned to extracellular regions and presented secretory pathway

232 signal peptides. DEG products identified only at ZT12 were related to the immune response,

233 amidation and kinase activity. Given the intermediate phenotype exhibited by C* flies, we

234 reanalysed the data, searching for DEGs where C* flies showed intermediate expression (D*

235 > C* >N* or N* > C* > D*; SI Appendix, Table S4). The list of DEGs consisted of 22 genes

236 at ZT0 and 62 at ZT12. Amongst the different functions represented by these new lists were

237 photoreception, circadian rhythm, sleep, Oxidation-reduction and mating behaviour were

238 over-represented in both D* and N* flies. For example, Rhodopsin 3 (Rh3) was up-regulated

239 in D* flies, while Rh2 and Photoreceptor dehydrogenase (Pdh) were down-regulated in N*

240 flies. pastrel (pst), a gene involved in learning and memory, was up-regulated in D* flies,

241 while genes involved in the immune response were up-regulated in N* flies. The only core

242 clock gene that showed differential expression was Par Domain Protein 1 (Pdp1), which was

243 up-regulated in D* flies. The clock output genes takeout (to) and pdf were up-regulated in N*

244 flies. Overall, the results suggested that genes that are transcriptionally associated with

245 diurnal preference are mostly found upstream (light input pathways), and downstream of

246 genes comprising the circadian clock.

247

248 Complementation test

249 We investigated the contribution of various genes to nocturnal/diurnal behaviour using a

250 modified version of the quantitative complementation test (QCT) (24). We tested the core

10 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

251 circadian clock genes per and Clk, the circadian photoreceptor cry and the output gene Pdf

252 and Pdfr, encoding its receptor (Fig. 9, SI Appendix, Table S5) (25). We also tested the ion

253 channel-encoding narrow abdomen (na) gene, given its role in the circadian response to light

254 and dark-light transition (26). The QCT revealed significant allele differences in per, Pdf,

255 Pdfr, cry and na (Fig. 9, SI Appendix, Table S5), indicative of genetic variability in these

256 genes contributing to the nocturnal/diurnal behaviour of the isogenic lines.

257 Since switching from nocturnal to diurnal behaviour in mice has been shown to be

258 associated with metabolic regulation (27, 28), we also tested Insulin-like peptide 6 (Ilp6), and

259 chico, both of which are involved in the Drosophila insulin pathway. A significant effect was

260 found in Ilp6 but not in chico (SI Appendix, S5). Other genes that failed to complement were

261 paralytic (para), encoding a sodium channel, and coracle (cora), involved in embryonic

262 morphogenesis (29-31).

263 We also tested genes that could affect the light input pathway, such as Arrestin2

264 (Arr2) and misshapen (msn) (32, 33). There was a significant evidence for msn failing to

265 complement but not for Arr2 (SI Appendix, Table S5). Various biological processes are

266 associated with msn, including glucose metabolism, as suggested by a recent study (34).

267

268

269 Discussion

270 In this study, we used artificial selection to generate two highly divergent populations that

271 respectively show diurnal and nocturnal activity profiles. The response to selection was

272 asymmetrical, as reflected by heritability h2, which was higher for diurnality (37.1%) than for

273 nocturnality (8.4%). This may indicate that different alleles and/or different genes were

274 affected in the two nocturnal/diurnal selections. Selections for traits affecting fitness have

11 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

275 been shown to have higher selection responses in the direction of lower fitness (19). This may

276 reflect the original (natural) allele frequency, whereby deleterious traits are mostly

277 represented by recessive alleles and favourable traits are represented by alleles at high

278 frequencies (19). This asymmetrical allele frequency could generate non-linear heritability,

279 such that a slower response to selection (as seen with nocturnal flies) is associated with

280 increased fitness (19). Indeed, nocturnal females lived longer and produced more progeny

281 than did diurnal females, an observation that supports a scenario of asymmetric

282 nocturnal/diurnal allele frequencies.

283 To what extent is the circadian clock involved in diurnal preference? We observed

284 that (i) PER cycling in the lateral neuron was significantly shifted in nocturnal flies, and (ii)

285 the phase of M and E peaks in DD differed between the strains, as did their free-running

286 period (particularly in the isogenic strains). On the other hand, our data indicate that a non-

287 circadian direct effect of light (light masking) played a significant role in diurnal preference,

288 particularly nocturnality, as nocturnal flies in DD conditions become rather diurnal (Fig. 5).

289 In rodents, the differential sensitivity of nocturnal and diurnal animals to light masking has

290 been well documented (35). This phenomenon was observed both in the laboratory and in the

291 field (36), with light decreasing arousal in nocturnal animals and the opposite effect occurring

292 in diurnal animals. Light masking in flies appears to have a greater impact, as it drives flies to

293 nocturnality.

294 As for the circadian clock, one can ask which part of the circadian system, if any, is

295 the target of selection (either natural or artificial) that shapes diurnal preference? Potentially,

296 the clock itself or the input (light) or output pathways (or a combination of these components)

297 could be targetted. Available evidence alludes to the latter being the case. First, the selection

298 lines generated here showed similar circadian periods and phases. In the isogenic strains (i.e.,

299 the N* and D* lines), however, there was a noticeable difference in the FRP, with a longer

12 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

300 period being seen in nocturnal flies. Indeed, long and short periods are expected to drive the

301 late and early chronotypes, respectively (37). The behavioural differences between the strains

302 were accompanied by 2-4 h changes in the phase of nuclear PER in the LNv (of note is the

303 fact that the sLNv receives light information from the eyes (38)). Most importantly, a

304 comparison of gene expression between diurnal and nocturnal flies highlighted just a single

305 core clock gene (pdp1). This finding is reminiscent of the results of our previous study in

306 flies (39), where transcriptional differences between the early and late chronotypes were

307 present in genes up- and downstream of the clock but not in the clock itself. The phase

308 conservation of core clock genes in diurnal and nocturnal animals has also been documented

309 in mammals (40-42). We thus suggest that selection for diurnal preference mainly targets

310 downstream genes, thereby allowing for phase changes in specific pathways, as changes in

311 core clock genes would have led to an overall phase change.

312 The main candidates responsible for diurnal preference that emerged from the current

313 study were output genes, such as pdf (and its receptor Pdfr) and to, as well as genes involved

314 in photoreception, such as Rh3, TotA, TotC (up-regulated in D* flies) and Rh2 and Pdh (up-

315 regulated in N*). Genes such as misshapen (msn) and cry were implicated by

316 complementation tests. RH3 absorb UV light (λmax, 347 nm) and is the rhodopsin expressed

317 in rhabdomer 7 (R7) flies (43, 44), while RH2 (λmax, 420 nm) is characteristic of the ocelli

318 (44, 45) and pdh is involved in chromophore metabolism (46).

319 The transcriptional differences between nocturnal and diurnal flies that we identified

320 are likely to be mediated by genetic variations in these genes or their transcriptional

321 regulators. Our current effort is to identify these genetic variations which underlie the

322 genetics of temporal niche preference. For this, the nocturnal and diurnal selection strains

323 generated here will be an indispensable resource.

324

13 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

325

326 Materials and Methods

327 SI Appendix has extended experimental details

328 Artificial Selection

329 To generate a highly genetically variable Drosophila melanogaster population, we pooled 5

330 fertilized females (4-5 days old) from 272 isofemale lines from 33 regions in Europe and

331 Africa (SI Appendix, Table S3) in the same culture bottle containing standard sugar food.

332 This population was maintained at 25°C in a 12:12 LD cycle. The progeny of this population

333 was used in the artificial selection as generation 0 (C0; Fig. 1). The locomotor activity of 300

334 males was recorder over 5 days in a 12:12 LD cycle at 25°C. Using the R library GeneCycle

335 and a custom-made script, we identified rhythmic flies and calculated their ND ratio (i.e., the

336 ratio between the amount of activity recorded during the night and during the day). For the

337 first cycle of selection, we selected 25 males with the most extreme nocturnal or diurnal ND

338 ratios, and crossed them with their (unselected) virgin sisters. The 10 following generations

339 underwent the same selection procedure. In addition, three control groups (CA, CB, CC)

340 were generated from the original population (C0) by collecting 25 fertilized females in 3 new

341 bottles. At each selection cycle, the controls underwent the same bottleneck as did the

342 nocturnal (N) and diurnal (D) populations but without any selective pressure. During and

343 following selection, the flies were maintained at 25°C in a 12:12 LD cycle.

344 Realize heritability was calculated for both the N and D populations from the

345 regression of the cumulative response to selection (as a difference from the original

346 population C0), with the cumulative selection differential being based on the data from the 10

347 cycles of selection (Fig. 1B) (47). Statistical differences between cycles of selection or

348 between lines were calculated using the Kolmogorov-Smirnov test (KS-test) and the Kruskal-

349 Wallis rank sum test.

14 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

350 To calculate the correlation between the ND ratios of parents and offspring, we

351 phenotyped 130 virgin males and 130 virgin females from the original population (C0) as

352 described above and randomly crossed them. We calculated ND ratios of the progeny of each

353 cross of a male and a virgin female and correlated this value with parental ND ratios.

354

355 Immunocytochemistry (ICC)

356 ICC was used to analyse the expression of PER and PDF in the fly brain. For quantification

357 of PDF levels, we used mouse monoclonal anti-PDF and polyclonal rabbit anti-PER

358 antibodies (48, 49). The protocol used for whole-brain staining was previously described (50,

359 51).

360

361 RNAseq

362 Gene expression was measured in head samples of flies collected at light-on (ZT0) and light-

363 off (ZT12) times. Flies (3-4 days old) were entrained for 3 days in 12:12 LD cycle at 25°C at

364 which point male heads were collected in liquid nitrogen at ZT0 and 12. RNA was extracted

365 using a Maxwell 16 MDx Research Instrument (Promega), combined with the Maxwell 16

366 Tissue Total RNA purification kit (AS1220, Promega). RNAseq library preparation and

367 sequencing was carried by Glasgow Polyomics using an IlluminaNextseq500 platform. Two

368 independent libraries (single-end) were generated per time point per line. The data were then

369 processed using Trimmomatic (version 0.32) (52) to remove adapters. The libraries were

370 quality checked using fastQC (version 0.11.2) (53). The total sequence obtained for each

371 library ranged from 9.7 Mbp to 21 Mbp, with a “per base quality score” > 30 phred and a

372 “mean per sequence quality score” > 33 phred. The RNA-seq was aligned to the Drosophila

373 melanogaster transcriptome (NCBI_build5.41) downloaded from the Illumina iGenome

374 website.

15 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

375

376

377 Figure legends

378 Fig. 1. Responses to artificial selection of Nocturnal/Diurnal locomotor activity. A.

379 Distribution of ND ratios of males of the starting population (C0, n=176). The insets show

380 actogram examples of diurnal (top) and nocturnal flies (bottom). Grey and yellow shading

381 represent night and day, respectively. B. Males average ND ratios of the selected populations

382 per selection cycle. The black solid line represents nocturnal selection, while the dashed line

383 represents diurnal selection. Data points correspond to average ND ratios ± standard

384 deviation (n=104-316). The ND ratio of the original population is shown in C0. C.

385 Correlation between mid-parents (n = 105) and mid-progenies (n =105). Correlation

386 coefficients and p values are reported below the regression equation. D. Correlation between

387 mothers (n =85) and daughters (n =85).

388

389 Fig. 2. Correlated responses of fitness traits to selection. A. Survival curves of flies of the

390 three selection populations (n=36-40; N: black, D: red, C: blue). The proportion of surviving

391 flies (y axes) is plotted against the number of days (x axes). B. Number of progeny produced

392 per female for N, C and D crosses. Grey and white bars indicate the number of males and

393 females, respectively (average ± standard error). Development time (egg-to-adult)

394 distributions are shown for male progeny (C) and females (D). Proportion (%) of progeny

395 produced per female per day. Males: N (n=3556), D (n=2308) and C (n=2998). Females: N

396 (n=3748), D (n=2401) and C (n=3145).

397

398 Fig. 3. Circadian behaviour of nocturnal and diurnal selection flies. A. Boxplots of

399 circadian periods under free-run conditions. The horizontal line in each box represents the

16 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

400 median value, the bottom and upper ends of the box correspond to the upper and lower

401 quartiles, respectively and the whiskers denote maximum and minimum values, excluding

402 outliers. Statistical differences were tested by a TukeyHSD test, with * signifying p < 0.05.

403 The FRP was longer in the C line (n =80) than in the D line (n=159), while the difference

404 between the D and N lines (n=240) was only marginally significant (TuskeyHSD test,

405 p=0.06). The FRP of N and C flies was similar (p =0.45). B. Acrophase angles of the free-run

406 activity for N (black circles, n=255), D (red circles, n=159) and C (blue triangles, n=65)

407 populations. The free-run phases of the three populations did not differ (F2,476:1.91, p =0.149,

408 NS). Lines represent mean vectors ± 95% confidence interval (CI). One hour corresponds to

409 an angle of 15°. C. Phase delay angles are shown for N (black circles, n =153), D (red circles,

410 n=155) and C (blue triangles, n=56) populations. Differences are not significant (F2,361 =1.47,

411 p =0.23, NS). Lines represent mean vectors ± 95% CI.

412

413 Fig. 4. Circadian behaviour of isogenic strains (N*, D* and C*). A. Boxplots of free-

414 running periods of the N* (n=64), C* (n=124) and D* (n=157) isogenic lines. Solid lines

415 represent median periods, the bottom and upper ends of the box correspond to the upper and

416 lower quartiles, respectively and the whiskers denote maximum and minimum values,

417 excluding outliers. TukeyHSD test, *p<0.05, **p<0.001 B. Acrophase angles of the free-

418 running activity shown for the N* (black circles, n=64), D* (red circles, n=157) and C* (blue

419 triangles, n=124) isogenic lines. The phase in the N* line was delayed by 2.02 h, as compared

420 to what was measured in the D* line, and by 1.38 h, as compared to what was measured in

421 the C* line (F2,342:6.01, p<0.01). Lines represent mean vectors ± 95% CI. One hour

422 corresponds to a 15° angle. C. Phase of eclosion in the N* (black circle, n=75), D* (red

423 circle, n=111) and C* (blue triangle, n=50) lines. The eclosion phase of D* flies was delayed

424 by ~2 h, as compared to what was measured with both the N* and C* lines (F2,233:4.95,

17 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

425 p<0.01). There was no difference between N* and C* flies (F1,123:0.08, p =0.78, NS). Lines

426 represent mean vectors ± 95% CI. One hour corresponds to an angle of 15°. Light-on (ZT0)

427 and light-off (ZT12) translated to 0° and 180° angles, respectively. D. Phase delays of N*

428 (black circles, n = 66), D* (red circles, n=170) and C* (blue triangles, n=126) flies. There

429 was no difference between the strains (F2,359:1.93, p = 0.15, NS). Lines represent mean

430 vectors ± 95% CI.

431

432 Fig. 5. Light masking of locomotor behaviour. A. Double plots of the median locomotor

433 activity (± SEM) per 30 min bin at 3 days in a 12:12 LD cycle followed by 4 days in DD.

434 (N*, black, n=71; D*, red, n=130; C*, Blue, n =110). Shades indicate light-off. Yellow

435 dashed lines delineate subjective nights. The overall ANOVA between 4 days in LD and 4

436 days in DD (starting from the second day in DD) indicates a significant effect of the light

437 regime (i.e., LD vs. DD; F1,594 =105.03, p<0.0001) and genotype (F2,594:173.01, p<0.0001).

438 The interaction genotype × environment was also significant (F2,594:102.66, p<0.0001). Post-

439 hoc analysis (TukeyHSD text) revealed that both N* and C* flies became significantly more

440 “diurnal” when released in constant conditions (N* p<0.0001; C* p<0.001). The ND ratios of

441 D* flies did not significantly change in DD (p=0.94, NS).

442

443 Fig. 6. Expression of PER in clock neurons. Quantification of nuclear PER staining in N*

444 (full lines) and D*(dashed lines) flies maintained in LD. Shaded area represents dark.

445 Representative staining (composite, Z-stacks) is shown below each panel. Points represent

446 averages ± standard error. The peak of PER signals in ventral neurons (LNv: 5th-sLNv,

447 sLNv, lLNv) was delayed in N* flies, as compared to D* flies (ZT1 vs ZT21-23), as

448 indicated by significant time x genotype interactions: 5th-sLNv χ2=12141, df=11, p<0.0001;

449 sLNv χ2=4779.4, df=11, p<0.0001; lLNv χ2=7858, df=11, p<0.0001. The N* PER signal is

18 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

450 stronger in sLNv (χ2=2416.6, df=1, p<0.0001), LNd (χ2=3924, df=1, p<0.0001) and DN1

451 (χ2=1799, df=1, p<0.0001) and weaker in DN2 (χ2=523.2, df=1, p<0.05).

452

453 Fig. 7. Expression of PDF in LNv projections. PDF staining in the N* (full lines) and D*

454 (dashed lines) lines maintained in a 12:12 LD cycle. Shading represents light-off.

455 Representative staining is shown in Supplementary Fig. S4. Points represent averages ±

456 standard error. The N* signal was lower than the D* signal at ZT3 (F1,18=11.99, p<0.01) and

457 ZT7 (F1,19=10.13, p<0.01), and higher at ZT11 (F1,15=10.53, p<0.01) and ZT13 (F1,17=23.39,

458 p<0.001).

459

460 Fig. 8. Functional annotation of DEGs associated with diurnal preference. Pie charts

461 representing significant terms of DEGs in 3 pairwise contrasts (D* vs N*, D* vs C*, N* vs

462 C*) at ZT0 and ZT12. Sections represent the percent of enrichment for each term. p<0.05

463 after Benjamini correction with the exception of the “signal peptide” term at ZT0, where p =

464 0.054.

465

466 Fig. 9. Quantitative complementation tests. Testing whether N*, D* and C* alleles vary in

467 terms of their ability to complement the phenotype caused by the Pdf and Pdfr mutant alleles.

468 Average ND ratio ± standard deviation of Pdf01 heterozygotes (left) and of Pdfr using p-

469 element insertions Pdfr5304 (middle) and Pdfr3369 (right). Numbers of tested flies are reported

470 in each chart bar. * represents p < 0.05.

19 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

471

472 Data Availability

473 The RNASeq sequencing files are available at the Gene Expression Omnibus (GEO)

474 accession GSE116985 (use reviewer token: ibspysiwzvafhef).

475

476 References

477 1. Hut RA, Kronfeld-Schor N, van der Vinne V & De la Iglesia H (2012) In search of a

478 temporal niche: Environmental factors. Prog Brain Res 199: 281-304.

479 2. Heesy CP & Hall MI (2010) The nocturnal bottleneck and the evolution of mammalian

480 vision. Brain Behav Evol 75(3): 195-203.

481 3. Daan S (1981) in Biological Rhythms, ed Aschoff J (Springer US, New York), pp 275-298.

482 4. Jacobs GH (2009) Evolution of colour vision in mammals. Philosophical Transactions of

483 the Royal Society B 364: 2957-2967.

484 5. Levy O, Dayan T & Kronfeld-Schor N (2007) The relationship between the golden spiny

485 mouse circadian system and its diurnal activity: An experimental field enclosures and

486 laboratory study. Chronobiol Int 24(4): 599-613.

487 6. Watson NF, Buchwald D & Harden KP (2013) A twin study of genetic influences on

488 diurnal preference and risk for alcohol use outcomes. J Clin Sleep Med 9(12): 1333-1339.

489 7. Archer SN, et al (2010) Polymorphism in the PER3 promoter associates with diurnal

490 preference and delayed sleep phase disorder. Sleep 33(5): 695-701.

20 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

491 8. Yoshii T, Rieger D & Helfrich-Forster C (2012) Two clocks in the brain: An update of the

492 morning and evening oscillator model in Drosophila. Prog Brain Res 199: 59-82.

493 9. Rieger D, et al (2007) The fruit fly Drosophila melanogaster favors dim light and times its

494 activity peaks to early dawn and late dusk. J Biol Rhythms 22(5): 387-399.

495 10. Kempinger L, Dittmann R, Rieger D & Helfrich-Forster C (2009) The nocturnal activity

496 of fruit flies exposed to artificial moonlight is partly caused by direct light effects on the

497 activity level that bypass the endogenous clock. Chronobiol Int 26(2): 151-166.

498 11. Bachleitner W, Kempinger L, Wulbeck C, Rieger D & Helfrich-Forster C (2007)

499 Moonlight shifts the endogenous clock of Drosophila melanogaster. Proc Natl Acad Sci U S

500 A 104(9): 3538-3543.

501 12. Fujii S, Krishnan P, Hardin P & Amrein H (2007) Nocturnal male sex drive in

502 Drosophila. Curr Biol 17(3): 244-251.

503 13. Tauber E, Roe H, Costa R, Hennessy JM & Kyriacou CP (2003) Temporal mating

504 isolation driven by a behavioral gene in Drosophila. Curr Biol 13(2): 140-5.

505 14. Kumar S, Vaze KM, Kumar D & Sharma VK (2006) Selection for early and late adult

506 emergence alters the rate of pre-adult development in Drosophila melanogaster. BMC Dev

507 Biol 6: 57.

508 15. Pegoraro M, et al (2014) Molecular evolution of a pervasive natural amino-acid

509 substitution in Drosophila cryptochrome. PLoS One 9(1): e86483.

21 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

510 16. Allada R, White NE, So WV, Hall JC & Rosbash M (1998) A mutant Drosophila

511 homolog of mammalian clock disrupts circadian rhythms and transcription of period and

512 timeless. Cell 93(5): 791-804.

513 17. Kumar S, Chen D & Sehgal A (2012) Dopamine acts through cryptochrome to promote

514 acute arousal in Drosophila. Genes Dev 26(11): 1224-1234.

515 18. Kim EY, et al (2002) Drosophila CLOCK protein is under posttranscriptional control and

516 influences light-induced activity. Neuron 34(1): 69-81.

517 19. Frankham R (1990) Are responses to artificial selection for reproductive fitness

518 characters consistently asymmetrical? . Genetic Research, Cambridge 56: 35-42.

519 20. Myers EM, Yu J & Sehgal A (2003) Circadian control of eclosion: Interaction between a

520 central and peripheral clock in Drosophila melanogaster. Current Biology 13(6): 526-33.

521 21. Qiu J & Hardin PE (1996) Developmental state and the circadian clock interact to

522 influence the timing of eclosion in Drosophila melanogaster. Journal of Biological Rhythms

523 11: 75-86.

524 22. Huang da W, Sherman BT & Lempicki RA (2009) Bioinformatics enrichment tools:

525 Paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res

526 37(1): 1-13.

527 23. Huang da W, Sherman BT & Lempicki RA (2009) Systematic and integrative analysis of

528 large gene lists using DAVID bioinformatics resources. Nat Protoc 4(1): 44-57.

22 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

529 24. Long AD, Mullaney SL, Mackay TF & Langley CH (1996) Genetic interactions between

530 naturally occurring alleles at quantitative trait loci and mutant alleles at candidate loci

531 affecting bristle number in Drosophila melanogaster. Genetics 144(4): 1497-1510.

532 25. Tataroglu O & Emery P (2015) The molecular ticks of the Drosophila circadian clock.

533 Curr Opin Insect Sci 7: 51-57.

534 26. Zhang L, et al (2010) DN1(p) circadian neurons coordinate acute light and PDF inputs to

535 produce robust daily behavior in Drosophila. Curr Biol 20(7): 591-599.

536 27. van der Vinne V, et al (2014) Cold and hunger induce diurnality in a nocturnal mammal.

537 Proc Natl Acad Sci U S A 111(42): 15256-15260.

538 28. van der Vinne V, Gorter JA, Riede SJ & Hut RA (2015) Diurnality as an energy-saving

539 strategy: Energetic consequences of temporal niche switching in small mammals. J Exp Biol

540 218(Pt 16): 2585-2593.

541 29. Peixoto AA & Hall JC (1998) Analysis of temperature-sensitive mutants reveals new

542 genes involved in the courtship song of Drosophila. Genetics 148(2): 827-838.

543 30. Schutte RJ, et al (2014) Knock-in model of dravet syndrome reveals a constitutive and

544 conditional reduction in sodium current. J Neurophysiol 112(4): 903-912.

545 31. Fehon RG, Dawson IA & Artavanis-Tsakonas S (1994) A Drosophila homologue of

546 membrane-skeleton protein 4.1 is associated with septate junctions and is encoded by the

547 coracle gene. Development 120(3): 545-557.

23 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

548 32. Lee SJ, Xu H, Kang LW, Amzel LM & Montell C (2003) Light adaptation through

549 phosphoinositide-regulated translocation of Drosophila visual arrestin. Neuron 39(1): 121-

550 132.

551 33. Houalla T, Hien Vuong D, Ruan W, Suter B & Rao Y (2005) The Ste20-like kinase

552 misshapen functions together with bicaudal-D and dynein in driving nuclear migration in the

553 developing Drosophila eye. Mech Dev 122(1): 97-108.

554 34. Ugrankar R, et al (2015) Drosophila glucome screening identifies Ck1alpha as a

555 regulator of mammalian glucose metabolism. Nat Commun 6: 7102.

556 35. Mrosovsky N (1999) Masking: History, definitions, and measurement. Chronobiol Int

557 16(4): 415-429.

558 36. Rotics S, Dayan T, Levy O & Kronfeld-Schor N (2011) Light masking in the field: An

559 experiment with nocturnal and diurnal spiny mice under semi-natural field conditions.

560 Chronobiol Int 28(1): 70-75.

561 37. Roenneberg T & Merrow M (2003) The network of time: Understanding the molecular

562 circadian system. Curr Biol 13(5): R198-207.

563 38. Schlichting M, et al (2016) A neural network underlying circadian entrainment and

564 photoperiodic adjustment of sleep and activity in Drosophila. J Neurosci 36(35): 9084-9096.

565 39. Pegoraro M, et al (2015) Gene expression associated with early and late chronotypes in

566 Drosophila melanogaster. Frontiers in Neurology 6: 100.

567 40. Ramanathan C, Stowie A, Smale L & Nunez AA (2010) Phase preference for the display

568 of activity is associated with the phase of extra-suprachiasmatic nucleus oscillators within

569 and between species. Neuroscience 170(3): 758-772.

570 41. Vosko AM, Hagenauer MH, Hummer DL & Lee TM (2009) Period gene expression in

571 the diurnal degu (Octodon degus) differs from the nocturnal laboratory rat (Rattus

572 norvegicus). Am J Physiol Regul Integr Comp Physiol 296(2): R353-61.

573 42. Otalora BB, Hagenauer MH, Rol MA, Madrid JA & Lee TM (2013) Period gene

574 expression in the brain of a dual-phasing rodent, the Octodon degus. J Biol Rhythms 28(4):

575 249-261.

576 43. Feiler R, et al (1992) Ectopic expression of ultraviolet-rhodopsins in the blue

577 photoreceptor cells of Drosophila: Visual physiology and photochemistry of transgenic

578 animals. J Neurosci 12(10): 3862-3868.

579 44. Stavenga DG & Arikawa K (2008) One rhodopsin per photoreceptor: Iro-C genes break

580 the rule. PLoS Biol 6(4): e115.

581 45. Feiler R, Harris WA, Kirschfeld K, Wehrhahn C & Zuker CS (1988) Targeted

582 misexpression of a Drosophila opsin gene leads to altered visual function. Nature 333(6175):

583 737-741.

584 46. Wang X, Wang T, Jiao Y, von Lintig J & Montell C (2010) Requirement for an

585 enzymatic visual cycle in Drosophila. Curr Biol 20(2): 93-102.

586 47. Falconer DS & Mackay TFC (1996) Introduction to Quantitative Genetics, (Longman,

587 Harlow).

25 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

588 48. Stanewsky R, et al (1997) Temporal and spatial expression patterns of transgenes

589 containing increasing amounts of the Drosophila clock gene period and a lacZ reporter:

590 Mapping elements of the PER protein involved in circadian cycling. J Neurosci 17(2): 676-

591 696.

592 49. Hermann-Luibl C, Yoshii T, Senthilan PR, Dircksen H & Helfrich-Forster C (2014) The

593 ion transport peptide is a new functional clock neuropeptide in the fruit fly Drosophila

594 melanogaster. J Neurosci 34(29): 9522-9536.

595 50. Hermann C, Yoshii T, Dusik V & Helfrich-Forster C (2012) Neuropeptide F

596 immunoreactive clock neurons modify evening locomotor activity and free-running period in

597 Drosophila melanogaster. J Comp Neurol 520(5): 970-987.

598 51. Hermann C, et al (2013) The circadian clock network in the brain of different Drosophila

599 species. J Comp Neurol 521(2): 367-388.

600 52. Bolger AM, Lohse M & Usadel B (2014) Trimmomatic: A flexible trimmer for illumina

601 sequence data. Bioinformatics 30(15): 2114-2120.

602 53. Andrews S (2010) FastQC: A quality control tool for high throughput sequence data.

603 available online at: Http://Www.bioinformatics.babraham.ac.uk/projects/fastqc

604

605

26 bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/380733; this version posted August 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.